Speech recognition techniques have been developed to recognize “what is being said” in a user utterance. For example, the speaker recognition techniques may be utilized to interact with an automated call center, input text, initiate commands, and so on. However, these techniques were typically developed for use regardless of what user provided the utterance. Accordingly, traditional speech recognition techniques were ill suited to identify a speaker, in other words, to determine “who” spoke.